Potential movement paths for male grizzly bear (Ursus arctos) dispersal between the Northern Continental Divide and Greater Yellowstone Ecosystems, 2000-2015

• Main Authors: Van Manen, Frank T., Peck, Christopher P., Costello, Cecily M., Haroldson, Mark A., Landenburger, Lisa, Roberts, Lori L., Bjornlie, Daniel D., Mace, Richard D.
• Format: Dataset
• Language: unknown
• Published: U.S. Geological Survey 2017
• Subjects: Grizzly Bears,; Ursus arctos
• Online Access: https://dx.doi.org/10.5066/f72v2f2w; https://www.sciencebase.gov/catalog/item/5911dbe2e4b0e541a03c1bac

For several decades, grizzly bear (Ursus arctos) populations in the Greater Yellowstone Ecosystem (GYE) and the Northern Continental Divide Ecosystem (NCDE) have increased in numbers and range extent. Whereas the NCDE population is contiguous with grizzly bear populations in the Canadian Rocky Mountains, genetic evidence suggests the GYE population remains isolated. Recent analyses indicate the effective population size of GYE grizzly bears has increased and is approaching levels needed for long-term viability. With only ~110 km distance separating current estimates of occupied range for these populations, the potential for immigration into the GYE from an NCDE migrant, or vice versa, is likely greater now than it has been for many decades. Therefore, we sought to delineate potential movement paths providing connectivity for male-mediated gene flow between the two populations. Using GPS locations for 124 male grizzly bears monitored during 199 bear-years, and covariates describing ecological, physical, and anthropogenic features of the landscape, we first developed step-selection functions to generate conductance layers of bear movements for each ecosystem. We then used randomized shortest path (RSP) algorithms to estimate the average number of net passages for all grid cells in the study region, when moving from an origin to a destination node. Given the conductance layer, movement paths follow certain grid cell sequences more than others and resulting RSP values thus provide a measure of movement potential. Repeating this process for 100 pairs of random origin and destination nodes, we identified movement paths for 3 levels of random deviation (Θ) from the least-cost path. We observed broad-scale concordance between model predictions for paths originating in the NCDE and those originating in the GYE for all 3 levels of movement exploration. Our model predictions indicate that grizzly bear movement between the ecosystems could involve a variety of routes, and verified observations of grizzly bears outside occupied range support this finding.